Frequency filtering, dip filtering, and gain control are three processes whose purposes seem to be largely cosmetic: they are used to improve the appearance of data. The criteria used to choose the quantitative parameters of these and similar processes are often vague and relate to human experience or visual perception. In principle, it should be possible to choose the parameters by invoking information theory and using objective criteria such as signal and noise dip spectra. But in routine practice this is not yet being done.
The importance of cosmetic processes is not to be underestimated.
On many occasions, for example, a comparison of processing techniques
(in order to choose a contractor perhaps?)
has been frustrated by an accidental change in cosmetic parameters.
These cosmetic processes arise naturally within wave-propagation theory.
It seems best to first understand how they arise,
and then to build them into the processing,
rather than try to append them in some artificial way after the processing.
The individual parts of the wave-extrapolation equations
will now be examined to show their cosmetic effects.